Sensitivity Analysis of CO2 Concentrations as Ventilation Metrics

An approach has previously been developed to estimate space-specific carbon dioxide (CO2) levels that can serve as metrics for the adequacy of outdoor ventilation rates. These metrics are based on the CO2 concentration expected in a space given its intended or expected ventilation rate, volume, and occupant information (i.e., the number of occupants, their CO2 generation rates, and duration of occupancy). This expected concentration can then be compared to a measured value to assess whether the actual outdoor air ventilation rate of the space is consistent with a design value, the requirement in a standard, or other recommended ventilation rate. A measured concentration higher than the expected value may indicate that the target ventilation rate is not being achieved. However, the occupant characteristics that impact the rate at which they generate CO2 (sex, age, body mass, and level of physical activity) are difficult to know with precision, which impacts the uncertainty in the expected concentration.
This study involved a sensitivity analysis of the calculated CO2 ventilation metric. Occupant characteristics impacting the CO2 generation rate (body mass, ratio of male-to-female occupants, and metabolic rates) were varied by about +/- 20 % to evaluate the impact on two metric values of interest (the CO2 concentration 1 hour after full occupancy and the steady state concentration). In addition, the space ventilation rate was varied by +/- 20 % to allow comparison to the variations associated with the other three inputs. The analysis employs the airflow and contaminant transport tool CONTAM to predict these concentrations over the range of inputs using factorial analysis. The sensitivity analysis employs a two-level full factorial design and a 10-step exploratory data approach (EDA) to identify the factors that have the most significant impacts. The result shows that outdoor ventilation rates and metabolic rates have the most significant effects on the CO2 ventilation metric values. Varying these two parameters by +/- 20 % results in variations in the CO2 concentrations of about 20 % to 35 %.